JPMorgan Research Analysis: AI Shopping Agents Go Live, Payments Still Routed Through Card Networks

By Rita

Six months ago, market expectations were that agency commerce would require complex merchant integration and agent agreement alignment. Now, Meta’s Muse pays non-integrated merchants using single-use virtual cards and integrated merchants using tokenized cards, making the payment flow more akin to traditional e-commerce. JPMorgan believes the order of the payment processing world is increasingly nearing normalcy, with limited near-term impact on covered companies.

In its October 23, 2026 industry report on payment processing, JPMorgan noted that Muse collaborates with Stripe Link, routing payments for non-integrated merchants through dual virtual card transactions and those for integrated merchants through tokenized card transactions. This mechanism resembles early non-integrated payment flows in food delivery and buy-now-pay-later sectors, generating dual transactions in the short term while relying on card-based infrastructure in the long term. Payment networks stand to benefit, and while online checkout faces competition, its trust and credit moats remain intact. The competitive landscape for traditional acquirers is better than previously expected, and the impact on new banks and remittance companies remains neutral.

The adoption speed of agency commerce, the dual-transaction mechanism of payment networks, the competitive dynamics of online checkout, and the shifting landscape for traditional acquirers are the four key dimensions for understanding recent stock trends in payments. Individual stock preferences are diverging, with overweight ratings concentrated on Affirm, Chime, Marqeta, Remitly, and Toast; PayPal and Klarna hold neutral ratings, while Western Union is rated underweight.

Agency Commerce Resembles Traditional E-Commerce More Closely

Understanding Muse’s payment flow is key to grasping the trajectory of agency commerce. For non-integrated merchants, payments are made via single-use virtual cards; Stripe Link issues these cards to consumers, which are then used to pay merchants, creating two card network transactions. For integrated merchants, payments draw directly from consumers’ underlying funding sources, primarily through tokenized card transactions.

Six months ago, market consensus held that agency commerce required complex merchant integration and agent contract alignment. Recent developments from Muse and Instinct indicate that browser-checkout models are becoming the standard for early-stage agency commerce. Browser checkout eliminates the need for merchant integration; agents scrape product information from webpages and complete checkout on behalf of users. While less reliable and efficient than headless checkout, it enables rapid deployment.

JPMorgan highlighted that allocation of transaction liability remains a core area of observation. Fraud and chargeback liability for card-present transactions defaults to the issuing bank, while card-not-present transactions default to the merchant. Liability for agency transactions has yet to be established, but adoption could accelerate if an entity is willing to act as a registered merchant and absorb the liability.

Card-Based Transactions Favor Payment Networks

The dual-transaction mechanism for non-integrated merchants provides short-term tailwinds for Visa and Mastercard. Most agency transactions generate two card network transactions and fees: one for funding the virtual card and another for paying the merchant via the virtual card. Transactions for integrated merchants are primarily tokenized card transactions, which will see increased use of network tokens over the long term.

Consumers strongly prefer funding agency transactions with cards or digital wallets. According to JPMorgan’s 2026 online checkout survey, only 3% of consumers prefer funding e-commerce transactions via bank accounts. Since agency transactions involve unknown counterparties, the fraud protection and dispute resolution rights offered by cards hold greater value.

JPMorgan maintains that the central role of payment networks in agency commerce remains unchanged. Short-term dual transactions represent a transitional revenue boost, while long-term tokenized transactions offer a more stable income stream. Agency commerce will not bypass traditional payment rails, and card-based infrastructure remains the core.

Online Checkout Moats Remain Intact

Agency commerce poses challenges to online checkout providers, with the most direct risk being the erosion of friction monetization. Consumers often select a specific checkout button out of convenience; if agents can absorb that friction, button usage rates may decline. However, JPMorgan notes that trust and credit moats persist.

The online checkout survey shows that approximately 30% of consumers are at least somewhat comfortable with agency transactions, and of those, around 50% prefer funding agency payments through wallets, with PayPal leading at a 32% preference rate. JPMorgan believes that digital wallets’ brand and trust advantages remain relevant in agency commerce, potentially intensifying wallet competition.

Replacing the credit moat proves more difficult. As intermediaries between consumers and merchants, agencies face higher underwriting complexity, giving technologically advanced providers a first-mover advantage. Affirm is well-positioned regarding underwriting complexity, fee structures, and consumer trust, while Klarna and PayPal leverage brand and trust advantages but must prove their ability to maintain market share within agency commerce.

Improving Landscape for Traditional Acquirers

With agency commerce trending closer to traditional e-commerce, the competitive environment for traditional acquirers has improved beyond six-month-ahead expectations. JPMorgan previously argued that the technical complexity of agency commerce would intensify e-commerce acquiring competition, favoring tech-savvy new payment service providers. The browser-checkout model reduces integration complexity, restoring importance to scale and cost advantages.

E-commerce acquiring has consolidated among five providers: Adyen, Braintree (PayPal), JPMorgan, Stripe, and Worldpay (Global Payments). Combined transaction volume for Stripe, JPMorgan, and Adyen increased by approximately $1 trillion last year, roughly half of the $2 trillion incremental volume added by Visa and Mastercard, representing a combined growth of about 20%.

Global Payments derives approximately 75% of its revenue from card-present transactions, insulating it from agency commerce acquiring competition. Fiserv has single-digit percentage revenue exposure to e-commerce acquiring, limiting the impact. PayPal’s Braintree holds a more favorable position in agency commerce, though it accounts for only low single-digit percentages of PayPal’s transaction processing revenue. Toast focuses on card-present verticals, minimizing agency commerce impact. Square’s card-not-present acquiring represents low-to-mid single-digit percentages of Block’s gross profit, keeping the impact manageable.

Limited Near-Term Volume Impact

The slow adoption rate of agency commerce is the key reason for limited near-term impact. Muse has recorded approximately 12 million global downloads, while Instinct boasts an annualized transaction volume of around $1 billion with hundreds of thousands of users. For Muse, 25% of users engage in shopping, with an average transaction value of $200.

Approximately 40% of Instinct users add a credit card within their first three weeks, though early adopters generally exhibit higher trust in AI than the general public. JPMorgan’s online checkout survey indicates that only about 30% of consumers are at least somewhat comfortable with agency commerce, a metric that better reflects near-term adoption likelihood. E-commerce currently accounts for 17.1% of US retail sales, having risen by roughly 9 percentage points over the past decade; shifting consumer spending habits takes time.

JPMorgan conducted sensitivity analyses on user counts and per-capita spending. Under a scenario of 60 million users spending an average of $6,000 annually, US agency commerce purchases would reach approximately $360 billion, representing 3.4% of Visa and Mastercard’s US purchase volume in 2026. Under a scenario of 100 million users spending $10,000 annually, the share rises to 9.3%. Given the small current user base and slow adoption rate, the near-term impact on covered companies remains limited.

Divergence in Individual Stock Preferences

JPMorgan favors Affirm for online checkout, citing its technical complexity and underwriting capabilities as well-suited for early integration into agency commerce. PayPal benefits from brand and trust advantages in wallet competition, while Klarna also possesses brand strength but must demonstrate its ability to retain market share.

In the new banking space, Chime offers low-cost banking services and stands to benefit as AI assistants help consumers find lower-fee banking options. JPMorgan views Chime’s 6% stock decline since Muse’s launch, underperforming the S&P 500 by roughly 7 percentage points, as an overreaction. Dave faces higher risks in liquidity product pricing, while Cash App’s instant deposit fees may face pressure, though swap and lending revenues remain unaffected.

Regarding remittances, both Remitly and Western Union price at mid-to-high levels relative to peers, a dynamic AI assistants could exacerbate through intensified price competition. JPMorgan identifies trust and payment network coverage as key moats; Remitly’s ability to gain market share despite higher prices underscores that trust and reliability outweigh price considerations. Remitly’s integration with ChatGPT improves the discovery process and could drive incremental transaction volume.

JPMorgan Coverage Ratings: Affirm (Overweight), Block (Overweight), Chime (Overweight), Dave (Overweight), FIS (Neutral), Fiserv (Neutral), Global Payments (Neutral), Klarna (Neutral), Marqeta (Overweight), PayPal (Neutral), Remitly (Overweight), Toast (Overweight), Western Union (Underweight).

The adoption pace of agency commerce, the allocation of transaction liability, and wallet competition will determine the next divergence point for payment stocks.

Disclaimer

This article is compiled and interpreted by TechFlow Research based on a third-party broker research report (JPMorgan, October 23, 2026), incorporating publicly available market information. The ratings, target prices, earnings forecasts, and related judgments cited herein reflect the views of the broker’s analysts, represent solely the stance of their respective institution, do not constitute the views of TechFlow Research, and should not be construed as investment advice.

Markets entail risk; investment decisions must be made independently. This article should not serve as the basis for buying or selling any securities.